04:19 <damo22> i am having trouble with LZMA ERROR with vendor uboot, so i disabled | lzma | stage in kernel wrappers, now i am getting it to tftpboot, but not with the vendor header (which is required for booting from flash to pass the kernel load address)

04:20 <damo22> my question is, how do i compute the KERNEL_LOADADDR ?

04:20 <damo22> and the KERNEL_ENTRY

04:20 <damo22> when theres no vendor header, it tftpboots just fine

04:22 <damo22> on my soc it tftpboots from 0x80000000

04:23 <damo22> but now theres a 0x200 byte header, and i tried 0x80000200 but that doesnt work

04:32 <damo22> going to try something

04:36 <dermoth> Hi, is there a reason ipsets in fw4 cannot be configured with "flags interval"? I want to use them for filtering with an external script updating the ipsets, but without an interval it's pretty moot for IPv6. I haven't seen this much but with normal delegation being between /48 and /64 bots could easily create one ip address per request and my ipset would fill up in seconds!

05:24 <owrt-images-builds> Build [#310](https://buildbot.openwrt.org/images/#/builders/182/builds/310) of `master_bmips/bcm6318` failed.

05:40 <dwfreed> dermoth: they can be, just not directly; you just set the match type to net, instead of ip

05:41 <dwfreed> dermoth: internally fw4 will set 'flags interval' for you in that case

05:42 <dwfreed> it will also automatically add 'auto-merge'

10:53 <KGB-2> https://tests.reproducible-builds.org/openwrt/openwrt_bcm47xx.html has been updated. (100.0% images and 100.0% packages reproducible in our current test framework.)

16:30 <dermoth> dwfreed: OK, I didn't realize that field was even being used, I though it only meant something if I use netfilter rules to update the set.

16:30 <dermoth> dwfreed: one detail though, updating the sets in-place didn't work (it did nothing, firewall reload from cli returned an error), I had to reboot the device to get the sets updated, although I could've probably deleted them but would have likely required deleting associated rules first...

17:21 <mrnuke> hurricos: I had some objections on https://github.com/Hurricos/realtek-poe/pull/31. Basically, it shouldn't be our job to support a JSON config format.

17:21 <mrnuke> hurricos: So I am expecting the submitter to drop the JSON config stuff from the series, so we can review de novo

17:23 <mrnuke> hurricos: https://github.com/Hurricos/realtek-poe/pull/36 will be painful to review. Sorry, but there isn't much I can do about that. The result should be easier going forward

19:48 <Mangix> Ansuel: ping

22:13 <russell--> weird, the netgear wndr3800 in my testbed decided it didn't have a wan on a recent build (r26938-192afd31e3) compared to my previous build (r26306-5ac8cf1eab)

22:14 <dwfreed> dermoth: some changes in fw4 require a full fw4 restart to rebuild the ruleset; reload just makes changes instead of completely rebuilding the ruleset it looks like

22:17 <dermoth> Ok... although the "restart" from the services page doesn't do that. I haven't looked much at the latest doc, maybe there is a simple way but I didn't see one in the web ui..

22:19 <dermoth> If I cared about affecting the traffic I wouldn't have done a fw restart anyway :)

22:40 <russell--> reconfirms that r26306-5ac8cf1eab works

23:12 <russell--> tried with a default config on 66177c081f1b, still no working eth1